Searches for Dark Matter with the ANTARES/KM3Net neutrino telescopes

Searches for Dark Matter with the ANTARES/KM3Net neutrino telescopes

S.R. Gozzini for the ANTARES and KM3NeT Collaborations

X CPAN DAYS Salamanca

29 Oct 2018

Overview

Neutrino telescopes have a wide scientific target

I neutrino astronomy

I dark matter

I multi-messengers

I particle physics

Telescopes in the Mediterranean

I field of view complementary to IceCube

I use water as a Cherenkov medium

ANTARES and KM3NeT ORCA

KM3NeT ARCA

KM3NeT ARCA and ORCA

Each of three building blocks: 115 strings, 64000 PMTs (31 PMTs / DOM and 18 DOMs / string)

ARCA (2 build. blocks)

string spacing: 90 m DOM spacing: 36 m Large sparse unit, high energies ORCA (1 build. block)

string spacing: 20 m DOM spacing: 9 m Small dense unit, low energies

Dark matter: indirect searches with neutrinos

Candidate: WIMPs, for example SUSY neutralino

I thermally produced in the early Universe

I relic density is blockedatfreeze-out

I mass ∼electroweak scale: ∼GeV<MWIMP <∼100 TeV Neutrino source in this case is a WIMP pair annihilation process

I can yield significant fluxes of high-energy ν

with SM =f¯f,W^±,qq¯

In short

Thermal bath (chemical + thermal equilibrium)

↓ Cooling Production stops

↓

Universe expansion Freeze-out: annihilation

stops

↓

Thermal equilibrium stops

DM particles created and annihilated in equilibrium with thermal bath

eg. γγ ↔DM DM

↓

Not enough energy to form DM mass No longer chemical equilibrium

↓

Volumes expand, number density decreases, annihilation rate decreases

Only elastic scattering

↓

DM in free range propagation

Sources

Relic WIMPs accumulate inmassive celestial bodieslike The galactic center

I highest signal expectation

I below horizon for detectors in Northern hemisphere

The sun

I sensitive to WIMP-nucleon cross-section (spin-dependent and spin-independent)

I clean signal, background well known

I less affected by halo uncertainties The Earth

Galaxy clusters

Signal: a cluster on the source

RA 150

− −100 −50 0 50 100 150

Dec

−80

−60

−40

−20 0 20 40 60

80 _{event_map}

Entries 605 Mean x −9.351 Mean y 37.34− RMS x 101.5 RMS y 20.76 event_map Entries 605 Mean x −9.351 Mean y 37.34− RMS x 101.5 RMS y 20.76 event_map_SIGNAL = 10

N

RA 150

− −100 −50 0 50 100 150

Dec

−80

−60

−40

−20 0 20 40 60

80 ^event_map

Entries 605 Mean x −13.46 Mean y −34.67 RMS x 101.7 RMS y 19.99 event_map Entries 605 Mean x −13.46 Mean y −34.67 RMS x 101.7 RMS y 19.99 event_mapSIGNAL = 100

N

Reproduced with MC simulations including a variable number of signal events, weighted according to a given DM model, over a number of background events taken from RA-shuffled data

Analysis Method

Unbinned likelihood analysis

logL(n_s) =

N

X

i=1

log [nsS(ψ_i,Ei,qi) +nbgB(δ_i,Ei,qi)]−nbg−ns

withS,Bdescribing the signal and background distribution of discriminating variables (angular informationψ, δ, energy estimate E, track reconstruction quality q).

Significance is computed comparing test statistics of data with distribution of pseudo-experiments with injected variablesignal

Signal

Most prominent channels yieldingν (assumed 100% B.R.) χχ¯→W⁺W⁻,bb, τ¯ ⁺τ⁻, µ⁺µ⁻, νν¯

dΦ_ν dEν

= 1 4π

hσvi 2M_χ²

dN_ν dEν

Z ∆Ω 0

dΩ Z

los

ρ²(r(s, θ, ψ))ds Observationn or limit µ₉₀ on integrated flux Φ =µ₉₀/(Acc·t)

µ90= hσvi 2

Z M

0

dN dEdE J

4π 1

M_χ² Acc(M_χ)t

number of events observed=annihilation rate * average number of particles per collision* source geometry * acceptance * time

Searches towards the Galactic Centre with ANTARES

dΦν

dE_ν = 1 4π

hσvi 2M_χ²

dNν

dE_ν Z ∆Ω

0

dΩ Z

los

ρ²(r(s, θ, ψ))ds Energy informationfrom the

PPPC tables[arXiv:1012.4515]

based on PYTHIA + oscillations

Morphology: J-Factor

I NFW, Einasto: cuspy, result from simulations

I Isothermal, Burkert: galactic rotation curves

Ingredients - Signal

Spatial: angular offset from GC drawing from J-Factor profile

Ψ smeared for source profile dP/dΩ Energy estimator (NHITS) : angular error estimateβ

Signal / Background

Data set: 2007-16 tracks (ν_µCC events)

PDF is built on energy distribution ofν and morphology Physical background (atmosphericν, mis-reconstructed atmosphericµ) are included in the likelihood

I spatial distribution: angular offset from source

I distribution of estimated energy

I reconstruction quality

Blind analysis: RA of real data is randomly shuffled until end.

Analysis procedure and results

Limits at 90% CL are set after finding no TS compatible with dark matter pseudoexperiment distribution in 10 years ANTARES data

I Sensitvity 90% CL means missing signal false negative less than 10% of the times

I Discovery 3σ means excludingfalse positiveless than 1-P(3σ)

PRELIMINARY

Limits on thermally-averaged annihilation cross-section

dΦν

dE_ν = 1 4π

hσvi 2M_χ²

dNν

dE_νJ

ANTARES different channels Results fromτ⁺τ⁻from different experiments

PRELIMINARY PRELIMINARY

Best limits forhigh WIMP masses: better angular resolution and higher effective volume (GC is in Southern hemisphere→ good visibility without veto)

Sensitivity estimated for KM3NeT

Promising chances - KM3NeT ARCA phase I (24 lines) 1 year

Searches towards the Sun

I Differential neutrino flux is related with the annihilation rate dΦ

dEν

= Γ

4πd² dN_ν dEν

I In equilibrium between capture and annihilation Γ =C/2 with C capture rate

I Flux only depends on WIMP-nucleon scattering cross section

I Very clean (BG well known); if signal→ direct interpretation

I Signal from moving source: bias-free

I Searches with neutrino telescopes crucial because sensitive at low velocities (= easier capture)

Searches towards the Sun

Sensitivities and limits on spin-dependent WIMP-nucleon cross section comparable with those from direct searches

Searches towards the Earth

WIMP capture and subsequent decay, but no equilibrium can be assumed due to low escape velocity

I WIMP scattering on Fe, Ni → spin-independent cross section

I Easier capture for WIMPs with mass ∼nucleus

I No easy background; non competitive with direct searches

Searches towards the Sun with secluded DM models

DM particlesecluded from SM matter by a mediator (GUT new gauge boson for instance). Possible signatures: double µ,direct annihilation intoν

10¹ 10² 10³ 10⁴

−5

−4

−3

−2

−1 0

M_χ [GeV]

log10[σSD (pb)]

ANTARES Di−µ; γcτ=1.5·10⁸ km ANTARES Di−µ; γcτ=2.8·10⁷ km ANTARES φ into ν; γcτ=1.5·10⁸ km ANTARES φ into ν; γcτ=2.8·10⁷ km PICO−2L

PICO−60 XENON100

10¹ 10² 10³ 10⁴

−10

−9

−8

−7

−6

−5

−4

−3

M_χ [GeV]

log10[σSI (pb)]

ANTARES Di−µ; γcτ=1.5·10⁸ km ANTARES Di−µ; γcτ=2.8·10⁷ km ANTARES φ into ν; γcτ=1.5·10⁸ km ANTARES φ into ν; γcτ=2.8·10⁷ km LUX 2013

XENON100 2012

Spin-dependentσ Spin-independent σ

Summary

I Search for dark matter profits from complementary methods:

indirect searches crucial

I Limits on cross-section for WIMP pair annihilation with 10 years ANTARES data. Best limits at high WIMP masses

I Limits for spin-dependent cross-section for WIMP-nucleon interaction, complementing with those from direct searches

I New scenarios can be tested: for instance secluded DM models. Wide range of possibilities in reach for KM3NeT.